Table 2

A binary logistic regression analysis of the association between household food insecurity and depressive symptoms among adolescent girls in north-eastern Ghana

Statistical modelFood secureMild food insecurityModerate food insecuritySevere food insecurity
OR (95% CI)P valueOR (95% CI)P valueOR (95% CI)P value
Model 1 (Crude)Ref1.54 (0.84 to 2.81)0.1602.40 (1.31 to 4.37)0.0053.22 (1.73 to 5.97)0.0002
Model 2Ref1.34 (0.72 to 2.49)0.3582.06 (1.09 to 3.87)0.0252.86 (1.48 to 5.53)0.002
Model 3Ref1.48 (0.78 to 2.81)0.2232.54 (1.31 to 4.92)0.0063.24 (1.65 to 6.38)0.001
Model 4Ref1.48 (0.78 to 2.80)0.2292.56 (1.32 to 4.96)0.0063.25 (1.65 to 6.40)0.001
Model 5Ref1.52 (0.80 to 2.89)0.2022.62 (1.35 to 5.13)0.0053.29 (1.66 to 6.49)0.001
Model 6Ref1.52 (0.80 to 2.90)0.2012.63 (1.35 to 5.12)0.0053.28 (1.66 to 6.49)0.001
  • Model 2 included child psychological factors including life satisfaction, self-efficacy, self-esteem and health complaints; model 3 additionally adjusted for other child-level factors including child’s age, menarche status, anaemia, stunting, mean frequency of consuming fruits and vegetables and mean frequency of animal foods consumption; model 4 was further adjusted for maternal age; model 5 was adjusted for household factors including wealth index and household size, and finally, model 6 adjusted for the intervention effect model fit statistic for final adjusted model: −2 Log L=520.08; Wald test=50.95, p<0.0001.